GMVar: Calculate standard deviation for Gaussian Mechanism

In difuture-lmu/dsROCGLM: ROC and calibration analysis for DataSHIELD

Calculate standard deviation for Gaussian Mechanism

Description

Calculate standard deviation for Gaussian Mechanism

Usage

GMVar(l2s, epsilon, delta, useAnalyticGM = True, ...)

Arguments

l2s	('numeric(1L)') l2-sensitivity.
epsilon	('numeric(1L)') First privacy parameter for (e,d)-differential privacy.
delta	('numeric(1L)') second privacy parameter for (e,d)-differential privacy.
useAnalyticGM	('logical(1L)') inicating whether noise for analytic Gaussian mechanism should be calculated. See details.

Details

If useAnalyticGM is TRUE, the standard deviation is calculated using the analytic Gaussian mechanism. The algorithm behind is a binary search which may cost a bit time, but the added noise is less. It also accepts epsilon > 1. If useAnalyticGM is FALSE, the standard deviation is calculated using the standard formula (faster, but a higher sigma).

Value

Numerical value for the standard deviation for the normal distribution.

Author(s)

Daniel S., Raphael R.

difuture-lmu/dsROCGLM documentation built on March 24, 2024, 1:07 p.m.